Hobbit regulates intracellular trafficking to drive insulin-dependent growth during Drosophila development

Abstract

All animals must coordinate growth rate and timing of maturation to reach the appropriate final size. Here, we describe hobbit, a novel and conserved gene identified in a forward genetic screen for Drosophila animals with small body size. hobbit is highly conserved throughout eukaryotes, but its function remains unknown. We demonstrate that hobbit mutant animals have systemic growth defects because they fail to secrete insulin. Other regulated secretion events also fail in hobbit mutant animals, including mucin-like 'glue' protein secretion from the larval salivary glands. hobbit mutant salivary glands produce glue-containing secretory granules that are reduced in size. Importantly, secretory granules in hobbit mutant cells lack essential membrane fusion machinery required for exocytosis, including Synaptotagmin 1 and the SNARE SNAP-24. These membrane fusion proteins instead accumulate inside enlarged late endosomes. Surprisingly, however, the Hobbit protein localizes to the endoplasmic reticulum. Our results suggest that Hobbit regulates a novel step in intracellular trafficking of membrane fusion proteins. Our studies also suggest that genetic control of body size, as a measure of insulin secretion, is a sensitive functional readout of the secretory machinery.

Keywords: Body size; Drosophila; Growth; Insulin secretion; Intracellular trafficking; Regulated exocytosis.

Fig. 1.

Identification of hobbit, a novel and conserved regulator of systemic growth. (A) hobbit mutant animals exhibit a dramatic small pupa phenotype. (B) ‘Tree of life’ phylogeny showing hobbit orthologs throughout Eukaryota, in green plants (green), animals (blue), fungi (red) and some protists (purple). (C) Protein identity plot showing primary sequence conservation between D. melanogaster hobbit and 12 metazoan orthologs (Caenorhabditis elegans, Bombyx mori, Anopheles gambiae, Danio rerio, Xenopus tropicalis, Gallus gallus, Oryctolagus cuniculus, Rattus norvegicus, Mus musculus, Macaca mulatta, Pan troglodytes and Homo sapiens); horizontal blue dashed line shows 100% identity. Vertical red dashed lines indicate the position of Drosophila hobbit mutant alleles, which are all nonsense mutations. hob: hobbit.

Fig. 2.

hobbit mutant animals can be rescued by fly and human proteins. Size quantification and lethal phase analysis for control, ubiquitous hobbit-RNAi [UAS-Dcr/+; tub>/UAS-hob(i) 1 and 2], hobbit mutants (hob²/hob³), hobbit mutant controls for rescue experiments (act>,hob²/hob³, hob²/UAS-hob,hob³ and UAS-Hs-hob, hob²/hob³), fly hobbit mutant rescue (act>,hob²/UAS-hob,hob³), and human ortholog rescue (UAS-Hs-hob,hob²/act>,hob³). Ubiquitous expression of hobbit-RNAi reduces body size and results in lethality during metamorphosis. Controls for rescue experiments do not significantly alter body size or lethality compared with hob²/hob³ mutant animals. Expression of fly hobbit or the human ortholog (Hs-hob) both rescue body size and lethality. Body size quantified by pupa volume is expressed as a percentage relative to wild type (100%). Data are shown as mean±s.e.m. Volume quantification and lethal phase analysis were performed on the same animals; n=100 animals per genotype. > in genotypes denotes GAL4; for example, ‘act>’ is shorthand for act-GAL4. A, adult; PA, pharate adult; P/iP, pupa/incomplete pupa; PP, prepupa.

Fig. 3.

hobbit has a non-cell-autonomous effect on growth. (A) Wing imaginal discs dissected from wandering L3 (wL3) larvae of control and hobbit mutant (hob¹/hob²) animals show that tissues are small in hobbit mutant animals. Nuclei stained with DAPI (white). (B) hob² mutant Flp/FRT clones (marked by loss of GFP and outlined in white) are the same size as paired twin spots (marked by 2× GFP and outlined in red) in both wing discs and salivary glands. Quantification of the ratio of clone:twin spot area (in arbitrary units, a.u.) confirms a 1:1 ratio between clone area:twin spot area. Data are shown as mean±s.d. n=10 clones/twin spots per tissue. (C) Analysis of developmental timing throughout larval development in control and hobbit mutant animals. Animals were synchronized at 0-4 h after hatching from the embryo (0-4 h after L1), then allowed to age until the onset of metamorphosis (puparium formation). hobbit mutant animals (dashed line) are delayed and asynchronous compared with control animals (solid line). y-axis shows percentage of animals pupariated, x-axis shows developmental time point in h after L1. Three replicates of n>50 animals (control total n=172, hob²/hob³ total n=214) were analyzed; data are shown as mean±s.d. (D) Analysis of developmental timing during the third larval instar in control and hobbit mutant animals. Animals were synchronized at 0-4 h after the start of the mid-third larval instar transition (0-4 h after Sgs3-GFP expression), then allowed to age until puparium formation. All control animals pupariate within 24 h of Sgs3-GFP expression (solid line); in contrast, hobbit mutant animals take considerably longer to pupariate (dashed line). y-axis shows percentage of animals pupariated, x-axis shows developmental time point in h after Sgs3-GFP expression. Three replicates of n>40 animals (control total n=146, hob²/hob³ total n=137) were analyzed; data are shown as mean±s.d. Scale bars: 50 µm.

Fig. 4.

hobbit is required for insulin secretion. (A) Immunofluorescence staining for Drosophila insulin-like peptide Dilp2 (red) in IPCs (marked with GFP in green) of control animals shows that Dilp2 is secreted under fed conditions, retained under starved conditions, and secreted within 2 h of re-feeding. (B) Dilp2 staining in hobbit mutant IPCs in fed, starved, and re-fed states shows that Dilp2 is not secreted under any condition. (C) IPC-specific hobbit RNAi knockdown (with Dilp2-GAL4) or fat body-specific hobbit RNAi knockdown (with Cg-GAL4) reduces body size. However, IPC- or fat body-specific (with ppl-GAL4 or Cg-GAL4) hobbit expression does not rescue hobbit mutant small pupae, whereas expression in both IPCs and fat body does rescue size. None of these treatments rescues hobbit mutant lethality. Size was quantified by pupa volume and expressed relative to wild type (100%); data are shown as mean±s.e.m. Volume and lethal phase data shown for hob²/UAS-hob,hob³ are the same as that shown in Fig. 2 but are included here for comparison. Volume measurements and lethal phase analysis were performed on the same animals; n=100 per genotype. ***P<0.0001 calculated by two-tailed t-test. (D) Representative image of body size rescue in hobbit mutant animals when hobbit is expressed in the IPCs only, fat body only, or IPCs and fat body. (E) Dilp2 (red) is not secreted in fed hobbit mutant animals upon expression of hobbit in the IPCs only (with Dilp2-GAL4) or fat body only (with ppl-GAL4). (F) Dilp2 (red) secretion is rescued in hobbit mutant animals when hobbit is expressed in both the IPCs (marked with GFP in green) and fat body (with Cg-GAL4). IPC images in all panels representative of n≥20 analyzed per condition/genotype. > in genotypes denotes GAL4; for example, ‘Dilp2>’ is shorthand for Dilp2-GAL4. A, adult; FB, fat body; n.s., not significant; PA, pharate adult; P/iP, pupa/incomplete pupa; PP, prepupa. Scale bars: 50 µm.

Fig. 5.

Secretory granules in hobbit mutant cells are not competent for exocytosis. (A) Flp/FRT hob² clone, marked by loss of RFP-nls (gray) and outlined in white, shows that Sgs3-GFP glue proteins (red) are not secreted in hobbit mutant salivary gland cells dissected from prepupal animals. (B) Sgs3-GFP glue granules in hobbit mutant salivary glands are significantly smaller than those in controls. Graph shows quantification by granule area; data are shown as mean±s.d. n≥60 granules per genotype. ***P<0.001 calculated by two-tailed t-test. (C) Synaptotagmin-1-GFP (Syt1-GFP, in green, expression driven in salivary glands by hs-GAL4), is loaded onto glue granule membranes in control glands but not in hobbit mutant glands. Glue protein shown in red. All images were acquired from live, unfixed tissue. Scale bars: 50 µm (A); 5 µm (B,C).

Fig. 6.

Trafficking of membrane fusion proteins on secretory granules is disrupted in hobbit mutant cells. (A) Syt1-GFP (green) localizes in large structures that are distinct from glue granules (red) in hobbit mutant cells. UAS-Syt1-GFP was expressed in salivary glands using hs-GAL4. Images were acquired from live, unfixed tissue. (B) Rab7-positive late endosomes (red) are dramatically enlarged in hobbit mutant salivary glands. Rab7-EYFP contains an EYFP tag at the endogenous Rab7 locus (Dunst et al., 2015). Images were acquired from live, unfixed tissue. (C) Rab7-positive endosomes (red) contain accumulations of the SNARE protein SNAP-24 (green) in hobbit mutant glands. (D) Hobbit-GFP (green, expression driven by Sgs3-GAL4) does not colocalize with Rab7 (red). Scale bars: 5 µm.

Fig. 7.

Hobbit localizes to the endoplasmic reticulum. (A) Hobbit-GFP (green) strongly colocalizes with the endoplasmic reticulum (ER) marker KDEL-RFP (red) in wandering L3 salivary glands. Expression of Hobbit-GFP and KDEL-RFP was driven by Sgs3-GAL4; images were acquired from live, unfixed tissue. (B) Hobbit-GFP (expressed using Sgs3-GAL4; green) does not appear to colocalize with the secretory granule membrane protein SNAP-24 (red). (C) Hobbit-GFP (green) does not colocalize with cytoplasmic mCherry (red) in glands at puparium formation (0 h PF). Expression was driven by Sgs3-GAL4; images were acquired from live, unfixed tissue. (D) KDEL-RFP (red) does not colocalize with cytoplasmic GFP (green) in glands at 0 h PF. Expression was driven by Sgs3-GAL4; images were acquired from live, unfixed tissue. (E) Hobbit-GFP, in green, strongly colocalizes with KDEL-RFP (red) in glands at 0 h PF. Expression was driven by Sgs3-GAL4; images were acquired from live, unfixed tissue. Scale bars: 50 µm (A); 5 µm (B-E).